VGA*-RRT*: A Mobile Robot Path Planning Algorithm for Irregular and Complex Maps
Minghao Duan, Zhou Wang, Xiyang Shao, Guangge Ren · IEEE Access · 2025
The aim of this paper is to propose a VGA*-RRT* planning model, which integrates the A* algorithm and the improved Rapidly Exploring Random Tree Star(RRT*) algorithm to address the path planning problem for mobile robots in irregular and complex environments. First, we present a practical obstacle map simplification strategy and utilize the A* algorithm to identify the optimal coarse path within the simplified grid maps. This step ensures that subsequent path generation remains consistently close to the actual optimal path. The optimal guidance vectors are then devised based on the direction vector information of the optimal path points. To reduce the sampling space, we define distinct sector sampling spaces, with the optimal guidance vectors serving as their median line reference. Additionally, we combine an improved vector field with the optimal guidance vector to create a new guidance vector that effectively directs sampling points away from local optima. After generating the initial path, multiple beta distribution points are inserted near the path points to enhance path optimization with higher probability. Finally, our adaptive rounded corner smoothing strategy enables effective rounding of corners while ensuring curvature continuity and producing a smooth overall path. To validate the effectiveness of the algorithm, we conducted simulation experiments using two irregular and complex maps with different levels of complexity. Compared to existing algorithms such as RRT*, Informed RRT*, and Sector Informed RRT*, the proposed method achieves approximately 20%-35% shorter path lengths across both map types while reducing planning time by approximately 70%-80%.